Legal safeguards in the age of Artificial Intelligence

Businesses are adopting Artificial Intelligence across operations, but legal experts warn companies must adopt clear policies, vendor controls, and contracts to protect data, ownership, and compliance.

As Artificial Intelligence reshapes business operations, legal advisers say companies must address new risks across recruitment, finance, cybersecurity, logistics, and customer service. Karen Painter Randall of Connell Foley LLP notes that Artificial Intelligence fosters innovation and efficiency while introducing “complex legal considerations” that require training, tools, and operational safeguards. Steven Teppler of Mandelbaum Barrett highlights implications for data governance, transparency, and protection of proprietary algorithms and training data as companies scale their use of these technologies.

Experts emphasize establishing a written Artificial Intelligence use policy and communicating it to employees. The article cites risks from employees using public tools such as ChatGPT and references the 2023 Samsung code leak as an example of inadvertent exposure of proprietary information. Nick Duston recommends that policies clarify what data may be exposed to Artificial Intelligence, set protocols for verifying AI-generated content to avoid hallucinations, require disclosure when content is AI-generated, and restrict use to vetted tools that mitigate data security risks.

Contract terms and ownership of inputs and outputs are central to mitigating risk. Wendi Uzar advises businesses to confirm that an Artificial Intelligence tool will not share a company’s information and to demand deletion of search queries and inputs within a specified timeframe. She also warns that AI-generated marketing or legal content can reproduce false or copyrighted material, and that website terms and conditions can create enforceable claims to protect material used to train models. Several pending district court cases are noted as creating uncertainty about outcomes.

Vendors and outside service providers are another focus. Teppler describes phone vendors using Artificial Intelligence to analyze voicemails, which can raise HIPAA or state-law concerns such as the New York Shield Act and New Jersey Data Privacy Act. Randall recommends multidisciplinary counsel-legal, business, and technology leaders-to identify blind spots and implement governance frameworks, risk assessments, bias audits, vendor contract reviews, incident response protocols, and employee education to balance innovation with protection.

50

Impact Score

Europe and US discuss biometric data-sharing framework

European Union and US officials are negotiating a border security arrangement that could enable continuous biometric data exchanges on EU citizens. The UK says the US has also requested access to fingerprint records as part of Visa Waiver Program discussions.

Apple plans Intel 18A-P for M7 and 14A for A21

Apple is expected to use Intel’s 18A-P process for M7 chips in MacBook models and Intel’s 14A process for A21 chips in iPhones. The shift points to a broader supplier strategy as Apple moves beyond TSMC for parts of its future silicon roadmap.

Google and other chatbots surface real phone numbers

Generative Artificial Intelligence chatbots are surfacing real phone numbers and other personal details, sometimes by pulling from obscure public sources and sometimes by inventing plausible but wrong contact information. Privacy experts say users have few reliable ways to find out whether their data is in model training sets or to force its removal.

U.S. and China revisit Artificial Intelligence emergency talks

Washington and Beijing are exploring renewed talks on an emergency communication channel for Artificial Intelligence as fears grow over the capabilities of Anthropic’s Mythos model. The shift reflects rising concern in both capitals that competitive pressure is outpacing safeguards.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.